A factorial design applied to polypropylene functionalization with maleic anhydride.
Polyolefin functionalization with polar monomers is a field that has been developed in recent decades, leading to new routes for obtaining advanced materials with improved technological properties. Reactions have been mainly carried out in solvent media (1-3), or in molten mixtures (4-7). New techniques, such as reactive extrusion and reactive injection molding, are being developed nowadays (8).
Important side reactions, such as chain scission and crosslinking, can occur during the polyolefin functionalization. When reactions are initialed by peroxide, it is well known that the dominant side reaction for polypropylene (PP) is chain scission (9, 10).
Maleic anhydride (MA) is the most commonly used polar monomer for polyolefin functionalization, possessing a high reactivity of the anhydride group in successive reactions and leading to thermally stable products (11). PP functionalization with MA has been extensively studied since the 1960s (12). Functionalized polymers have been used for the compatibilization of immiscible polymer blends as well as for the improvement of interfacial adhesion in polymer composites.
Despite the large number of published studies, there is no unanimity about certain mechanistic features. The exact site of attachment and the structure of the MA grafts have been investigated. Working with atactic PP in a solution process, Minoura (12) proposed that grafting reactions involve the appendage of MA single units to tertiary carbons along the backbone. Later, other authors (9, 13, 14) suggested that MA single units, as well as short blocks, could be appended to PP tertiary carbon atoms and to chain ends after chain scission reactions.
A simplified reaction mechanism is shown in Scheme 1. The generation of radicals is accompanied by degradation and crosslinking (recombination). Chain scission is expected at a lower radical concentration due to monomolecular degradation reactions. Crosslinking reactions start to prevail at higher radical concentrations due to their bimolecular character (15). Steric factors are important in preventing the crosslinking reactions.
In spite of the large number of studies dealing with PP functionalization with MA, a more profound and systematic investigation should be carried out, permitting not only the evaluation of parameters that affect these reactions but also evaluating possible synergistic and antagonistic interaction effects. Since such a study involves a prohibitively large number of experiments, chemometric procedures based on multivariate statistical techniques can be used.
In this work, a factorial design was performed to determine the statistical effects of MA concentration, DCP concentration, temperature, and time on the degree of functionalization and molecular weight of PP. In the factorial design, two levels of each variable were defined.
Commercial polypropylene (PPH301), without additives and a melt flow index 12g/10 min (2.16kg), Mn: 54,000, Mw/Mn: 4.95, was kindly supplied by OPP Petroquimica, Triunfo-RS and it was used as received. Maleic anhydride 99.5% (Produtos Quimicos Elekeiroz S.A.) was ground to a powder prior to use.
Dicumyl peroxide (DCP) 98% and dibenzoyl peroxide (DBP) 70% from Aldrich Chemical Company were used as purchased.
Commercial xylene and acetone were distilled prior to use.
The reactions were performed in a mixer chamber Haake Rheomex 600p, heated initially to 170 [degrees] C and saturated with argon, PP was introduced and after two minutes a mixture of MA and DCP was added. The speed of both rotors was fixed at 50 rpm. After a fixed period of time the product was removed, cut in small pieces, and extracted with acetone to remove unreacted MA and other volatiles.
Fourier-Transform Infrared Spectroscopy (FTIR)
An FTIR spectrometer Mattson 3020, Galaxy 3000 series was utilized for infrared analysis. Measurements were made using films prepared at 170 [degrees] C, 2.5 ton/[cm.sup.2], in a Carver press, Monarch Series, model 3710-ASTM. The scanned wavenumber range was 4000-400 [cm.sup.-1]. The absorbance of the carbonyl stretching of the grafted MA at 1784 [cm.sup.-1] was compared with the absorbance of the methyl groups at 1156 [cm.sup.-1] (A2 - internal reference) in order to evaluate the degree of functionalization.
Gel Permeation Chromatography (GPC)
Molecular weights were determined by means of gel permeation chromatography (GPC) on a Waters 150 CV system equipped with three columns, Styragel HT3, HT4, and HT6 ([10.sup.]3, [10.sup.4], and [10.sup.6]) and a refractive index detector. Analyses were undertaken using 1,2,4-trichlorobenzene as a solvent, at 140 [degrees] C, and the molecular weights were calculated using universal calibration curve constructed with 20 polystyrene. three polypropylene and three polyethylene standards.
Determination of MA on Modified PP
The amount of the grafted anhydride was determined by titrating the acid groups after complete hydrolysis [TABULAR DATA FOR TABLE 1 OMITTED] of the anhydride groups. In a typical procedure. 0.3 - 0.5g of functionalized PP was dissolved in 80 ml of water saturated xylene at 100 [degrees] C. The system was kept refluxing for one hour, then it was hot titrated with ethanolic KOH 0.02N, using phenolphthalein as an indicator,
RESULTS AND DISCUSSION
The first set of experiments included 16 reactions, and two levels were employed for the following parameters: maleic anhydride concentration [MA] (1.5 and 2.5 wt%), dicumyl peroxide concentration (DCP) (0.5 and 1.0 wt%), temperature [170 and 180 [degrees] C), and reaction time (10 and 20 minutes). Table 1 shows schematically this experimental set.
Reactions performed in the mixer were torque monitored. Typical torque curves are represented in Fig. 1. These values are indirectly related with molecular weight and, thus, function as a rough indication for degradation or crosslinking. It becomes evident that, under these conditions, the products' torques were always smaller than for pure processed PP. This is an indication of chain scission reactions. The highest [MA] shows a small torque indicating a small degree of degradation for this experimental set,
The FTIR spectra of the initial PP and of a typical grafted PP are represented in Fig. 2, A strong absorption band is observed for the grafted polymer at 1785 [cm.sup.-1], corresponding to the carbonyl stretching (A1). This absorption is compared with the absorption at 1156 [cm.sup.-1] (A2) characteristic of the PP methyl group. Ratios A1/A2 can be related to MA incorporated on the PP chains.
The plot of functionality (tool %, obtained by titration) against A1/A2, (absorbance ratios from FTIR) shows a linear relation, as demonstrated in Fig. 3. This graph has been used as a calibration curve for the products in order to determine the absolute functionality, Functionality values showed in Table 1 were obtained by interpolating FTIR ratios (A1/A2) in this plot,
Figure 4 shows the influence of the parameters and their interaction effects on functionality (F), torque, and Mw. Torque and Mw were directly correlated, as expected. Both of them increased with increasing [MA] suggesting that MA plays an important role on inhibiting chain-breaking reactions. Surprisingly, under the conditions employed, [MA] resulted in a negative effect on degree of functionalization, i.e., the highest [MA] resulted in a smaller functionalization, It appears that at high [MA] many competitive reactions take place that do not lead to grafting,
Considering as isolated factors, increasing [DCP] or temperature produce increasing functionality. This is easy to explain because both factors result in a greater number of free radicals in the medium, improving the grafting reaction. However, both factors show negative effects on torque and molecular weight. This means that higher free radical concentrations produce higher degradation, as expected,
As a result of the statistical analysis, it can be concluded that time does not affect the degree of functionalization but possesses a small negative effect on torque and Mw.
On the other hand, there is an important second order interaction effect between MA and temperature. This effect is positive on functionalization, although the main effect of [MA] is negative. The same interaction effect is negative on torque and Mw, but in both cases lower than the isolated effects [MA] and temperature. Thus at 180 [degrees] C and [MA] = 2.5% the system reactivity is higher, with more graft and degradation reactions compared with the reactions at 170 [degrees] C and [MA] = 1.5%. Figures 5 and 6 illustrate this second-order effect (nonlinear behavior) on torque and Mw assuming [DCP] = 1% and time = ten minutes. The similarity of these figures denotes the correlation between torque and Mw, as expected. From these plots [ILLUSTRATION FOR FIGURE 4 OMITTED] it is possible to conclude that degradation diminishes with decreasing temperature and increasing [MA]. The effect of temperature shows little influence on chain scission for low [MA].
Figures 7 and 8 show the degrees of functionality as a function of [DCP] and [MA] concentrations for 170 and 180 [degrees] C, respectively. It becomes obvious that temperature plays an important role in these reactions. At 170 [degrees] C, higher functionalization occurs for higher [DCP] and lower [MA]. The generation of radical sites is reduced at this temperature and increasing [MA] [TABULAR DATA FOR TABLE 2 OMITTED] does not enhance its incorporation. At 180 [degrees] C, functionalization is maximized for high levels of [DCP] and [MA]. In this case, temperature and [DCP] provided a larger number of radical sites, permitting more MA molecules to incorporate on the polymer chain. It is important to observe that high [DCP] and low [MA] produce almost the same functionalization for both temperatures.
Since the best values of the degrees of functionalization in the first design have been obtained at 180 [degrees] C and ten minutes, these values were chosen as constants for a new set of experiments, aiming to explore other [MA]. [DCP] levels were maintained, since increasing its concentration could result in higher degradation. Six new experiments were developed. The conditions and obtained results are summarized in Table 2.
The results presented in Fig. 4, show that the reaction time has no effect on functionality (F), Mw and torque. Then, reactions with 20 minutes can be considered as a replication of the experiments at ten minutes. This allows for the analysis of data from Tables 1 and 2 together, drawing response surfaces with respect to F, torque, and Mw as a function of [MA] and [DCP]. Considering the reaction time as ten minutes and a temperature of 180 [degrees] C, these surfaces are illustrated in Figs. 9 and 10.
For large values of F, an optimum value of [MA] between 1.4 and 2.6 wt% [ILLUSTRATION FOR FIGURE 9 OMITTED] has been found. Functionality also increases with [DCP] within this [MA] range. On the other side, Mw shows a minimum value when [MA] varies within this range [ILLUSTRATION FOR FIGURE 10 OMITTED]. The shape of both response surfaces is very similar but they are inverted, suggesting that high functionalization is accomplished by chain scission.
The initial increase in functionality with [MA] might be explained by the increasing probability of reactions between a macromolecular radical and an MA molecule. At higher [MA], the monomer can reach a limit of solubility in the polyolefin that would induce phase separation and decrease functionalization. A part of the initiator might remain at the MA phase and prevent chain scission reactions.
The performed experimental design shows the main effects for the analyzed parameters [MA], [DCP], reaction time, and temperature. Also, interaction effects between parameters have been determined.
The degree of functionalization and the molecular weights depend on [MA], [DCP], and temperature, Reaction times have no significant effect, at least not within the reaction times adopted in the present study. Interaction effects between [MA] and temperature play an important role on both the degree of functionalization and molecular weights.
At 180 [degrees] C, higher degrees of functionalization could be obtained with an [MA] between 1.4 and 2.6 wt%. Functionalization also increases with [DCP]. Increasing the degree of functionalization was directly related to the decrease in molecular weight, suggesting that maleic anhydride incorporation occurs mainly at the chain ends, after or before chain scission reactions.
The authors thank PADCT-NM, CNPq, and CAPES for financial support.
1. S. N. Sathe, G. S. S. Rao, and S. Devi, J. Appl. Polym. Sci., 53, 329 (1994).
2. G. de Vito, N. Lanzetta, G. Maglio, M. Malinconico, P. Musto, and R. Palumbo, J. Polym. Sci.; Polym. Chem. Ed., 22, 1335 (1984).
3. R. Greco, G. Maglio, P. Musto, and G. Scarinzi, J. Appl. Polym. Sci., 37, 777 (1989).
4. N. G. Gaylord, M. Mehta, and R. Mehta, J. Appl. Polym. Sci., 33, 2549 (1987).
5. M. Aglietto, R. Bertani, G. Ruggeeri, and F. Ciardelli, Makromol. Chem., 193, 179 (1992).
6. R. M. Ho and C. H. Wu, Polym. Prepr., 33(1), 941 (1992),
7. K. J. Ganzeveld and L. P. B. M. Janssen, Polym. Eng. Sci., 32, 467 (1992).
8. T. Vivier and M. Xanthos, J. Appl. Polym. Sci., 33, 2513 (1987).
9. N. G. Gaylord and M. K. Mishra, J. Polym. Sci.,: Polym. Lett. Ed., 21, 23 (1983).
10. R. M. Ho, A. C. Su, C. H. Wu, and S. Chen, Polymer, 34(15), 3264 (1993).
11. R. Greco, G. Maglio, and P. V. Musto, J. Appl. Polym. Sci., 33, 2513 (1987).
12. Y. Minoura, M. Ueda, S. Mizunuma, and M. Oba. J. Appl. Polym. Sci., 13, 1625 (1969).
14. B. de Roover, J. Devaux, and R. Legras, J. Polym. Sci.: Part A: Polym. Chem., 34, 1195 (1996).
15. E. Borsig, A, Fiedlerova, and M. Lazar, J. Macromol. Sci. Chem., A16(2), 513 (1981).
|Printer friendly Cite/link Email Feedback|
|Author:||Nachtigall, S.M.B.; Neto, R. Baumhardt; Mauler, R.S.|
|Publication:||Polymer Engineering and Science|
|Date:||Apr 1, 1999|
|Previous Article:||Cycle time predictions for the rotational molding process with and without mold/part separation.|
|Next Article:||Thermal and mechanical properties of the precursor polymers: comparison of their properties with poly(amic acid).|